Evolution and AI

I just finished reading Ishmael, one of the more thought-provoking books I’ve read in months. At one point in the book, during their discussions about evolution and man’s “creation myth”, Ishmael tells the main character a story:

This story (Ishmael said) takes place half a billion years ago—an inconceivably long time ago, when this planet would be all but unrecognizable to you. Nothing at all stirred on the land, except the wind and the dust. Not a single blade of grass waved in the wind, not a single cricket chirped, not a single bird soared in the sky. All these things were tens of millions of years in the future. Even the seas were eerily still and silent, for the vertebrates too were tens of millions of years away in the future.

But of course there was an anthropologist on hand. What sort of world would it be without an anthropologist? He was, however, a very depressed and disillusioned anthropologist, for he’d been everywhere on the planet looking for someone to interview, and every tape in his knapsack was as blank as the sky. But one day as he was moping along beside the ocean he saw what seemed to be a living creature in the shallows off shore. It was nothing to brag about, just a sort of squishy blob, but it was the only prospect he’d seen in all his journeys, so he waded out to where it was bobbing in the waves.

He greeted the creature politely and was greeted in kind, and soon the two of them were good friends. The anthropologist explained as well as he could that he was a student of life-styles and customs, and begged his new friend for information of this sort, which was readily forthcoming. “And now,” he said at last, “I’d like to get on tape in your own words some of the stories you tell among yourselves.”

“Stories?” the other asked.

“You know, like your creation myth, if you have one.”

“What is a creation myth?” the creature asked.

“Oh, you know,” the anthropologist replied, “the fanciful tale you tell your children about the origins of the world.”

Well, at this, the creature drew itself up indignantly—at least as well as a squishy blob can do—and replied that his people had no such fanciful tale.

“You have no account of creation then?”

“Certainly we have an account of creation,” the other snapped. “But it is definitely not a myth.”

“Oh, certainly not,” the anthropologist said, remembering his training at last. “I’ll be terribly grateful if you share it with me.”

“Very well,” the creature said. “But I want you to understand that, like you, we are a strictly rational people, who accept nothing that is not based on observation, logic, and the scientific method.”

“Of course, of course,” the anthropologist agreed.

So at last the creature began its story. “The universe,” it said, “was born a long, long time ago, perhaps ten or fifteen billion years ago. Our own solar system—this star, this planet and all the others seem to have come into being some two or three billion years ago. For a long time, nothing whatever lived here. But then, after a billion years or so, life appeared.”

“Excuse me,” the anthropologist said. “You say that life appeared. Where did that happen, according to your myth—I mean, according to your scientific account.”

The creature seemed baffled by the question and turned a pale lavender. “Do you mean in what precise spot?”

“No. I mean, did this happen on the land or in the sea?”

“Land?” the other asked. “What is land?”

“Oh, you know,” he said, waving toward the shore, “the expanse of dirt and rocks that begins over there.”

The creature turned a deeper shade of lavender and said, “I can’t imagine what you’re gibbering about. The dirt and rocks over there are simply the lip of the vast bowl that holds the sea.”

“Oh yes,” the anthropologist said, “I see what you mean. Quite. Go on.”

“Very well,” the other said. “For many millions of centuries the life of the world was merely microorganisms floating helplessly in a chemical broth. But little by little, more complex forms appeared: single-celled creatures, slimes, algae, polyps, and so on.

“But finally,” the creature said, turning quite pink with pride as he came to the climax of his story, “but finally jellyfish appeared!”

The story parallels our culture’s “creation myth” – our story of how creation and evolution inevitably lead to modern humans. This creation myth assumes humanity is evolution’s final output; the pinnacle of a process that’s been occurring for tens of billions of years.

I think this view changes within the next 30 years.

From an evolution standpoint, our greatest achievement may be the creation of super-intelligent AI. A billion years from now, what will be more important – the development and reign of mankind (for 11,000 years of civilization), or the subsequent rise and development of a super-intelligent AI?

No technology has as much potential to change the human condition as artificial intelligence. AI is improving at an astounding rate. You have computer programs driving cars more safely than humans. You have programs that can solve CAPTCHAs, do a better job diagnosing medical conditions than trained physicians, build realistic virtual realities, fly planes autonomously (or play ping pong), and trade the stock market. What’s more, the processing power and capacity for these programs to learn is increasing every day.

Given how rapidly AI is improving, and how important it’s becoming to our economy and way of life, it’s inevitable that it continues to take a larger role in society. Almost as inevitable is the fact that we will develop a superintelligent AI that is actually smarter than human intelligence. One that can help solve some of the many complex problems our economy, governments and businesses face.

My guess is that this shift happens gradually. Smart businessmen buy a program that determines the best strategic move for a given business. Next, computer scientists at a prestigious university release software that predicts (with stunning accuracy) the effects of economic and political policies on the nation. As these programs are able to model the impact of certain policies and suggest optimal courses of action, more and more policy decisions are placed in the hands of these superintelligent programs.

AI will change a lot, but it’s biggest impact may be on the way humanity perceives itself. What will it mean when humanity is no longer the one charting our destination as a species, but instead putting decisions in the hands of superintelligent AI? What’s more, once we reach this state there is no rational reason to take back control from AI – why would you place immense decisions about economic policy in the hands of a human when a much better decision can be made by AI?

Right now, humanity is the jellyfish. Our culture believes that we’re the end result of creation and evolution – everything up to this point was guided towards the creation of our species. Within the next century we’ll be in a different position. Ultimately, I think modern humanity will be looked at the same way we now look at Homo Erectus – as an important evolutionary step in the creation of the real pinnacle of evolution: superintelligent AI.

0